【摘要】 机器人视觉中 ,对多面体物体图像的线条解释技术一直沿用Huffman的线条解释方法。通过对Huffman方法的研究 ,指出了其中的缺陷。例如 ,它无法识别某些类型的简单物体 ;在给线条作标记时亦不具备逻辑上的统一性 ,使得有时出现信息冗余 ,有时又出现信息不足。该文提出克服缺陷弥补不足的改进办法 ,在标记面临几种选择时引入了优先级的概念 ,并对Huffman线条模型进行了调整。更多还原

【Abstract】 By Now,the main technology for line interpreting of polyhedra image has been the Huffman method.By testing Huffman method of line interpreting,its undoubted shortcomings are discovered:It cannot distinguish some kinds of objects, and its mark system has some interior contradictions.These problems are solved by making some changes to Huffman mark model and by developing a priority-rank model for mark choosing.更多还原